Good morning,
I am running a biprobit model in Stata 14, and want to stop the first two "fitting" equations after a max number of iterations. (In order to skip over non-converging models... I am running a bootstrap, and occasionally a sample will not converge.) However, the option iter(#) controls only the iterations run by the last, full MLE model. Here is a toy example --- the full MLE model is stopped at 5 iterations, but the first two models are not.
clear
sysuse auto
gen doms=foreign==0
gen forn=foreign==1
gl m1cov headroom mpg rep78 weight displacement gear_ratio
gl m2cov headroom length rep78 weight displacement headroom trunk
biprobit (doms $m1cov) (forn $m2cov), iter(5)
Is there any way to specify the biprobit command such that I can control max iterations for the 1st two models? Something along the lines of "biprobit (doms $m1cov, iter(5)) (forn $m2cov, iter(5))"?
Thanks,
Leah
I am running a biprobit model in Stata 14, and want to stop the first two "fitting" equations after a max number of iterations. (In order to skip over non-converging models... I am running a bootstrap, and occasionally a sample will not converge.) However, the option iter(#) controls only the iterations run by the last, full MLE model. Here is a toy example --- the full MLE model is stopped at 5 iterations, but the first two models are not.
clear
sysuse auto
gen doms=foreign==0
gen forn=foreign==1
gl m1cov headroom mpg rep78 weight displacement gear_ratio
gl m2cov headroom length rep78 weight displacement headroom trunk
biprobit (doms $m1cov) (forn $m2cov), iter(5)
Is there any way to specify the biprobit command such that I can control max iterations for the 1st two models? Something along the lines of "biprobit (doms $m1cov, iter(5)) (forn $m2cov, iter(5))"?
Thanks,
Leah